A systematic review identified tactile assessments used in children with cerebral palsy (CP), but their reproducibility is unknown. Sixteen children with unilateral CP and 31 typically developing children (TDC) were assessed 2-4 weeks apart. Test-retest percent agreements within one point for children with unilateral CP (and TDC) were Semmes-Weinstein monofilaments: 75% (90%); single-point localization: 69% (97%); static two-point discrimination: 93% (97%); and moving two-point discrimination: 87% (97%). Test-retest reliability for registration and unilateral spatial tactile perception tests was high in children with CP (intraclass correlation coefficient [ICC] = 0.79-0.96). Two tests demonstrated a learning effect for children with CP, double simultaneous and tactile texture perception. Stereognosis had a ceiling effect for TDC (ICC = 0) and variability for children with CP (% exact agreement = 47%-50%). The Semmes-Weinstein monofilaments, single-point localization, and both static and moving two-point discrimination are recommended for use in practice and research. Although recommended to provide a comprehensive assessment, the measures of double simultaneous, stereognosis, and tactile texture perception may not be responsive to change over time in children with unilateral CP.